EMNLP 2025LLM ReasoningSelf-Reflection
J. Li, Y. Zhou, J. Lu, G. Tyen, L. Gui, C. Aloisi and Y. He
A dual-model reasoning framework within a verbal reinforcement learning paradigm, decoupling inference-time self-reflection into specialised models for reasoning critique and refinement.
View PDFEMNLP 2025PersonalisationFeedback
R. Zhao, A. Bobrov, J. Li and Y. He
A modular LLM-based system that generates personalised, curriculum-aligned feedback in science education with error-aware assessment and an educator-in-the-loop interface.
View PDFEMNLP 2025AssessmentExplainability
J. Li, A. Bobrov, R. Zhao, C. Aloisi and Y. He
An interactive visualisation platform using multiple LLMs to score student answers and generate explanatory rationales with intuitive annotation tools.
View PDFACL 2025FaithfulnessInference
J. Li, H. Yan and Y. He
A probabilistic inference paradigm using task-specific and lookahead rewards to make LLM-generated rationales more faithful to model decisions and input context.
View PDFAAAI 2025AssessmentLLM
J. Li, A. Bobrov, D. West, C. Aloisi and Y. He
An automated and explainable educational assessment system for interactive and visual evaluation of student responses.
View PDFEMNLP 2024ReasoningScience
J. Li, H. Xu, Z. Sun, Y. Zhou, D. West, C. Aloisi and Y. He
A framework that improves scoring by generating thought trees and using synthetic data for supervised fine-tuning and preference optimisation.
View PDFEMNLP 2023DistillationAssessment
J. Li, L. Gui, Y. Zhou, D. West, C. Aloisi and Y. He
A framework for concurrent student-answer scoring and rationale generation that improves performance while preserving explainability.
View PDF